Martin J. Bayly

The ‘re-turn’ to empire in IR: colonial knowledge communities and the construction of the idea of the Afghan polity, 1809-38 Article (Accepted version) (Refereed)

Original citation: Bayly, Martin J. (2013) The ‘re-turn’ to empire in IR: colonial knowledge communities and the construction of the idea of the Afghan polity, 1809-38. Review of International Studies, 40 (3). pp. 443-464. ISSN 1469-9044 DOI: http://dx.doi.org/10.1017/S0260210513000338 © 2013 British International Studies Association This version available at: http://eprints.lse.ac.uk/63654/ Available in LSE Research Online: September 2015 LSE has developed LSE Research Online so that users may access research output of the School. Copyright © and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in LSE Research Online to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain. You may freely distribute the URL (http://eprints.lse.ac.uk) of the LSE Research Online website. This document is the author’s final accepted version of the journal article. There may be differences between this version and the published version. You are advised to consult the publisher’s version if you wish to cite from it.

1

The ‘re-turn’ to empire in IR: Colonial knowledge communities and the construction of

the idea of the Afghan polity, 1809-38

Martin J. Bayly

Author Bio:

Martin Bayly is a PhD candidate at the Department of War Studies, King’s College

London, and a Teaching Fellow at the Defence Studies Department, King’s College

London. The author wishes to thank Theo Farrell, Nicholas Onuf, Chris Bayly, Mervyn

Frost, Hameed Shuja, Chris Sims and three anonymous reviewers for their very helpful

comments on earlier drafts. Thanks also to the participants at the International Political

Sociology Doctoral Workshop 2012, especially Natália Maria Félix de Souza, as well as

the participants at the ‘Wars in Afghanistan’ panel at the International Studies

Association Annual Convention 2013. All errors and omissions are the sole

responsibility of the author. The PhD research upon which this paper draws was funded

by a studentship from the Defence Studies Department, King’s College London.

Institutional Address:

Department of War Studies, 6th

Floor, King’s College London, Strand, London, WC2R

2LS

Abstract:

This article seeks to add to the exploration and development of Imperial History’s

contribution to the discipline of International Relations (IR). Focusing on British

perceptions of Afghanistan in the period preceding the first Anglo-Afghan war the

article considers colonial knowledge as a source of identity construction, but in a

manner that avoids deploying anachronistic concepts, in this case that of the Afghan

‘state’. This approach, which draws on the insights brought to IR by historical

sociology, shows that engaging with Imperial History within IR can encourage a more

reflexive attitude to core disciplinary categories. This not only reveals alternative

approaches to the construction of specific political communities but it also allows for a

more historicist mode in the use of history by IR as a discipline. Furthermore, by

moving away from material based purely on diplomatic history, Afghanistan’s imperial

2

encounter can be recovered from the dominance of ‘Great Game’ narratives, offering an

account that is more appreciative of the Afghanistan context.

In January 2002, barely two months after Operation Enduring Freedom had toppled the

Taliban regime in Afghanistan, the former soldier, diplomat, and now Conservative MP

Rory Stewart walked across Afghanistan, following in the footsteps of the Mogul

Emperor Babur the Great. The title he chose for the work that followed the journey was

The Places In Between,i a reflection perhaps, of the enduring tendency to view

Afghanistan as at the confluence of the ‘knowable’ – of civilizations, empires, nation-

states, or societies – and yet paradoxically resembling a land of the ‘unknowable’, of

‘wild tribes’, nefarious actors; a domain of rumour, intrigue, and violence.ii As a former

history tutor at Balliol College Oxford, Stewart was certainly aware that by making his

trip he was also following in the footsteps of a collection of nineteenth century

explorers - themselves often scholars, soldiers, and diplomats - all seeking to uncover

this mysterious location ‘in between’ the imperial interests of British India, and Imperial

Russia.iii

As with Stewart, these individuals produced works, and in some cases,

provided the advice for policy makers as they sought to devise an effective foreign

policy on India’s northwest frontier.iv

The significance of this nineteenth-century colonial quest for knowledge of

Afghanistan, and its relevance to the political decisions that were taken, has often been

overlooked as a result of the more dramatic story of Anglo-Russian rivalry over Central

Asia; the so-called ‘Great Game’. Commonly this quest is simply portrayed as a story of

the swashbuckling adventures of a nascent colonial intelligence community, combating

the spread of Russian influence.v But as Benjamin Hopkins has shown, this was ‘far

from the only, nor even the most important’ story at this time.vi

The problem here is not

one of fact, but of emphasis.

In this article I aim to make two main contributions. Firstly, I seek to add to the

development, and the exploration of, imperial history’s contribution to the discipline of

International Relations (IR). In particular the article considers colonial knowledge as a

source of identity construction for the colonial state, and therefore providing a line of

3

enquiry for contemporary IR theorists. But I seek to do so in a manner that avoids

deploying anachronistic concepts, in this case that of the Afghan ‘state’.vii

This

approach, which draws on the insights brought to IR by historical sociology, shows that

engaging with imperial history within IR can encourage a more reflexive attitude to core

disciplinary categories. This not only offers insights into the construction of specific

political communities but it also allows for a more historicist mode in the use of history

by IR as a discipline. The need for this approach within IR is an argument frequently

made, but less frequently carried out.viii

Secondly, and to this end, I seek to contribute to the growing literature that seeks to

recover Afghanistan’s imperial encounter.ix

The common refrain that the British knew

nothing of Afghanistan prior to the First Anglo-Afghan war needs revision. But more

importantly, the article seeks to provide an alternative narrative to the traditional focus

on geopolitics and grand strategy – a bias which is itself a reflection of the tendency for

international history to focus on diplomatic sources to the detriment of more local or ‘on

the ground’ accounts.x By drawing on a sociology of knowledge approach, the article

charts the emergence of a set of criteria that provided a degree of legibility for policy-

makers, and in turn helped to order their policy prescriptions. Based on the observation

that knowledge is in part a ‘participation in the cultural resources of society’xi

this

provides a more cultural basis for foreign policy decision making, but one which is

inherently tied up in the process of interaction: of knowledge ‘becoming’ so, rather than

simply ‘being’.xii

The analysis focuses on the activities, correspondence and texts of a select group of

itinerant explorers, military men, archaeologists, adventurers, and quasi-official East

India Company representatives. Whilst initially uncoordinated, these individuals, I

argue, began to resemble a ‘knowledge community’ through their correspondence and

through the pooling of their work by an increasingly interested policy elite. Ultimately

this knowledge community would contribute to the decision to launch an invasion of

Afghanistan in 1838 in order to depose the ruler Dost Muhammad Khan for the

preferred former-King Shah Shuja ul-Mulk. This community therefore created the

groundwork for the emergence of the idea of Afghanistan in the imperial mind, as well

as the foundations for policy prescriptions, and would leave an indelible mark on British

conceptualizations of the Afghan strategic space.

4

Imperial history, IR, and historical sociology

History and International Relations, it has been noted, have not always enjoyed an easy

relationship.xiii

In the early 2000s there was a refocusing on this ongoing schism.xiv

On

the one hand it was argued that such a debate would help to overcome certain ahistorical

attitudes that had plagued IR throughout much of the Cold War period, tendencies that

were in part a result of the transhistorical commitments inherent to the dominant

rationalist and positivist theoretical approaches.xv

In a good example of this critique,

George Lawson noted IR’s tendency to present history as ‘scripture’: ‘the mining of the

past in order to confirm suppositions about the present; the smoothing out of

differences, varieties, and processes of change in the interests of methodological purity

and theoretical rigidity; and the bracketing off of history behind an eternal ‘‘illusory

present’’’.xvi

In short, IR was not just ahistorical, but also ahistoricist; reluctant to

engage reflectively with categories such as ‘state’, ‘system’, ‘sovereignty’, ‘power’ and

‘empire’.xvii

As these scholars pointed out, a more historically sensitive IR would uncover the

nuanced nature of such categories, demonstrating their historically contingent meanings,

rather than black boxing them as essentialised, even reified categories. The coincidental

rise of critical theory and constructivist approaches with their emphasis on linguistics,

subjectivity, and the continual reproduction of social institutions offered a more

questioning attitude to core categories and ontologies.xviii

Yet for some, this had only

gone so far. As Colin Wight argued, ‘rather than embarking on new theoretical or

empirical avenues, many scholars merely “poured the newly emerging patterns of

thought into the old framework”’.xix

The ‘cultural turn’ that constructivist approaches

had carried into the field remained, for some, lacking in emancipatory spirit, tied as they

often were to simple identity-based binary narratives of enemy/friend, or wedded to

familiar (state-based) objects such as ideas of national strategic cultures. The prospect

that constructivism – with its intellectual heritage deriving in part from critical theory –

would offer a more historicist approach in IR’s engagement with history has in some

respects fallen short of expectations.xx

5

This ‘historical turn’ (or ‘re-turn’ as Lawson rightly labels it)xxi

entailed a move away

from IR’s preoccupation with diplomatic history, which was becoming increasingly

marginalized within disciplinary history anyhow, and a turn towards social and cultural

history. Some called for greater attention in particular to imperial history,xxii

a field that

had itself undergone a ‘cultural turn’, yet one by which IR remained curiously

unmoved. As Tarak Barkawi, one of the leading proponents of an ‘imperial turn’,

pointed out, ‘[r]epeatedly, it would seem, IR was founded amidst empire, but

discovered instead only a world of sovereign states and their collective action

problems.’ As he notes, the failure of social science and IR to deal with questions of

empire and imperialism left the discipline inadequate ‘to the experiences and histories

of most of the peoples and places on the planet’.xxiii

For Barkawi and Laffey,

engagement with imperial history offered a threefold benefit of escaping the ‘territorial

trap’,xxiv

highlighting the importance of hierarchy in international relations, and

engaging with international relations as ‘thick’ social, political, cultural, and military

exchange. Central to this was the proposal that histories of the European and non-

European world were co-implicated in each other; that imperialism highlighted process

of co-constitution in state identity.xxv

This ‘imperial turn’ has yielded some fruitful returns. Jordan Branch has shown how the

spread of the territorial state was not simply the exporting of Eurocentric constructs of

sovereign authority, but rather a process of ‘colonial reflection’ whereby colonial

officials unfamiliar with local spatialities of power and authority were forced to rely on

a more intelligible ‘scientific’ approach through cartography – a method that was then

imported back to Europe.xxvi

Edward Keene, through his reframing of Grotius’ work has

highlighted how the concept of ‘divided sovereignty’ – the idea that sovereignty could

be shared by two powers - provided theoretical justification for imperialist ventures

throughout the nineteenth century.xxvii

Meanwhile Gerry Simpson - although not

focusing exclusively on empire - has nonetheless shown how during the nineteenth

century imperial entities were responsible for the construction of a legal framework that

institutionalized a form of ‘legal hegemony’ and ‘anti-pluralism’; one that mandated

distinctions between ‘Great Powers’ and ‘outlaw states’, a distinction that he argues has

left a legacy to this day.xxviii

Imperial history provides a reminder of the historical

contingency of core categories whose ontological stability is often taken for granted.

Equally, it shows that these categories were continually contested, not just within

6

imperial political thought, but in the process of global exchange, of which imperialism

was simply one variant. In short, the constitutive effects of imperial exchange are often

overlooked by an IR discipline that remains analytically imprisoned by its own theory-

driven orthodoxies.

The documenting of Afghanistan’s imperial encounters provides good examples of the

dangers of inadequate engagement with international history. Accounts often exhibit a

‘continuist myth’ by presenting Afghanistan as the perennial location for competition

between great powers and as the ‘graveyard of empires’. In addition, as Rob Johnson

has argued, Afghanistan’s history is frequently looted, or ‘contested’, in order to

provide a policy science for today’s challenges, drawing ‘a legacy of half-understood

and often misconceived ideas from a long period of colonial contact, and distant

memories of the Mujahideen struggle against Soviet Occupation in the 1980s’.xxix

An

important outcome of both of these trends has been the suppression of what could be

described as the ‘Afghanistan context’. As Thomas Barfield eloquently puts it, ‘[a]ll the

focus on war and visiting conquerors overshadows the country’s own inhabitants,

except as the rough warriors who served as speed bumps on the highway of conquest or

more recently earned a reputation for making the place ungovernable. As a result,

Afghanistan itself remains just the vague backdrop in a long-running international

drama where others hold the speaking parts.’xxx

In the case of the first Anglo-Afghan War, the status of Afghanistan is often suppressed

for a wider concern with Anglo-Russian relations – the so-called ‘Great Game’ – an

example of diplomatic history par excellence. The idea that the British were attempting

to establish a ‘buffer state’ regularly fails to engage with what exactly was meant by a

‘state’ in this context. Meanwhile policy failures are often attributed to generic and

theoretically reversible administrative or political blunders rather than dealt with on

their own terms.xxxi

The extent to which Afghanistan emerged in part as a figment of the

British imagination, and the co-implication of local agents, including Afghans

themselves in the crafting of their own nationality under the shadow of imperial power

lacks attention. That this should be so not only highlights a lacuna in the historiography

of this part of the world, but also offers insights in to the faint legacies that remain to

this day concerning representations of Afghanistan itself, as well as representations of

its political heritage – a heritage that is profoundly international.

7

As a way out of the inadequacies of IR’s engagement with imperial history in general,

and as a contribution to the recovery of Afghanistan’s imperial encounter in particular, I

propose closer attention to the knowledge community that emerged concerning policies

on the East India Company’s northwest boundary in the lead up to the First Anglo-

Afghan war. The reasoning here is that rather than assuming the Afghan state as an

ontological prior, we can use a more localized approach, drawing on the sociology of

knowledge, to uncover the calling into being of the Afghan polity as an actor; in this

case, a process that was carried out by outsiders. Simply casting back over the archive

in search of a nascent ‘state’ forms risks the charge of deploying an anachronism - in

this case relocating a historically contingent social institution into a historical context in

which such a category carried a different meaning.xxxii

Avoiding this requires attention

to the language used, and a more historically sensitive approach.

The fact is that the British rarely, if ever, referred to the concept of the Afghan ‘state’.

Indeed, the lack of Afghan ‘statishness’ provided in part a justification for the First

Anglo-Afghan War. The polity they envisaged was a conceptually heterogeneous entity.

It encompassed ethnographic definitions based around British taxonomies of the tribal

groups they had documented, and principal lineages within these. But it was also a

geographic expression, delineated by highland and lowland groups, bounded by

geographic features such as the Hindu Kush, and combining principal population

centres: specifically Kabul, Kandahar, and Herat. Finally, patterns of authority were

also multiple and often appeared in competition, thereby complicating a simple

presumption of any single sovereign authority. Alongside the expressions of ‘King’,

‘Sirdar’, ‘Khan’, and ‘tribe’ was the practice of the assigning of land tenures in return

for military service, which provided a more measurable form of authority. The

ambiguous status of non-Afghan groups, including diasporic merchant communities

(Hindus, Armenians, Jews), and nomadic groups added to the confusion. And finally

there was the less frequently noted role of Islam. The Afghan polity could therefore be

described as an essentially contested concept, however, the knowledge community that

formed around the definition of Afghanistan, and the concomitant growth of imperial

policy interest forced a policy of ‘closure’ around this concept.xxxiii

8

Although the purpose of this article is not to innovate within the well-developed

literature on the sociology of knowledge, it is necessary to begin with a brief outline of

how I conceive this approach to be helpful in building the argument. Firstly,

notwithstanding the considerable philosophical debate over what ‘knowledge’ really is,

for the purposes of this article, and borrowing from Peter Burke, I understand

knowledge to mean information that has been ‘‘cooked’, processed or systematized by

thought.’xxxiv

Noting Durkheim’s dictum that first one must understand social facts as

things, Berger and Luckmann’s seminal work, The Social Construction of Reality, gives

the practical suggestion that ‘common-sense ‘knowledge’ rather than ‘ideas’ must be

the central focus for the sociology of knowledge’xxxv

Whilst I agree with these

sentiments, the role of ‘thought’ in the systematization of information, does make it

necessary to refer to the deeper intellectual currents - often in the form that could be

loosely referred to as ‘ideas’ - in order to contextualize that which passes for

knowledge. In this sense, knowledge is in part a ‘participation in the cultural resources

of society’.xxxvi

Secondly, and related to this point, the forms of knowledge referred to in this article are

not conceived of as resembling a purely intellectual or academic exercise. Colonial

knowledge in particular provides capacities for action as a via-media between ideas and

action.xxxvii

As Nicholas Dirks has pointed out concerning Bernard Cohn’s work: ‘The

colonial state [can be] seen as a theater for state experimentation, where historiography,

documentation, certification, and representation were all state modalities that

transformed knowledge into power’.xxxviii

Indeed, Cohn’s ‘investigative modalities:

particularly those described as the ‘observational/travel’; ‘survey’; and ‘historiographic’

mode, provide a helpful backdrop in terms of contextualizing colonial knowledge

practices at this time.xxxix

Although Afghanistan was never directly colonized, through a

growth in colonial knowledge, the country was rendered more legible – albeit partially -

for policy-makers. The representations that were included within the repertoire of this

knowledge order were not necessarily recorded for the purposes of manipulation, yet

through an implicit and sometimes explicit value attachment, they carried at least the

potential for manipulation, should it be required, converting un-annexed territory into a

more familiar ‘realm of possibility’.xl

Indeed, as James C. Scott has highlighted, such

quests for legibility can be seen as a general inclination of states, not just the colonial

9

state, showing that such a process in itself can lead to a tenuous reflection of reality by

default. As he argues:

‘Certain forms of knowledge and control require a narrowing of vision. The

great advantage of such tunnel vision is that it brings into sharp focus certain

limited aspects of an otherwise far more complex and unwieldy reality. This

very simplification, in turn, makes the phenomenon at the center of the field of

vision more legible and hence more susceptible to careful measurement and

calculation. Combined with similar observations, an overall, aggregate synoptic

view of a selective reality is achieved, making possible a high degree of

schematic knowledge, control, and manipulation.’xli

The knowledge community leading up to the First Anglo-Afghan War resembles a

proto-knowledge order that provided the foundations for this later enduring aggregate

synoptic view of Afghanistan.

Thirdly, despite the apparent and actual projection of culturally-rooted, intellectual

trends, and representations onto the Afghan space, this form of knowledge capture and

sequestering should not be misinterpreted as a one-way process from metropole to

periphery, or as an imperialist imposition of Eurocentric categories onto a non-

European space. This was certainly part of the story, but as more recent work (at least

since the late 1990s) on this aspect of imperial history has shown has shown, imperial

expansion, including practices of knowledge procurement, also involved exchange and

interaction both within and between imperial territories, peripheries, and elsewhere.xlii

An important aspect of British knowledge on Afghanistan was accordingly derived in

part from their experience elsewhere in South Asia and Persia.xliii

Moreover, their

reliance on local informants, including Afghan agents, as well as members of diasporic

communities within Kabul and elsewhere presented a picture that derived from more

than a purely Eurocentric bias.xliv

Information, information, information: the emergence of a colonial knowledge

community

10

The Afghanistan ‘knowledge community’ in the early stages, and right up to the first

Anglo-Afghan war was a highly select group. Hopkins puts the documented number of

Europeans visitors to Afghanistan prior to 1838 at just fifteen.xlv

Of particular relevance

were those who filled an official capacity subsequent to their journeys. Three

individuals stand out: Firstly, Mountstuart Elphinstone, who was the first envoy to the

Court of Kabul in 1809, later becoming the Governor of Bombay.xlvi

Key to

Elphinstone’s influence was his 1815 two-volume published text, An Account of the

Kingdom of Caubul, which was the most significant outcome of his 1809 mission.xlvii

A

second key figure was Charles Masson, a deserter from the Bengal Army,xlviii

who later

became a newswriter and agent of the British in Kabul from 1834-8. Finally, Alexander

Burnes, whose two trips under the auspices of the East India Company (EIC) led to his

appointment as envoy to Kabul during the First Anglo-Afghan War. Alongside this core

group we can also identify a more peripheral group of actors including Henry Pottinger,

whose 1810 trip skirting the southern and western reaches of Afghanistan was followed

later with a posting as political agent at Hyderabad during the 1830s;xlix

William

Moorcroft, who in 1819 set off across Kashmir through Kabul and to Balkh; and finally

Claude Wade, who recruited Masson, and as Political Agent in Ludhiana filled an

important role processing and filtering intelligence reports.l

There was an entrepreneurial spirit to these ventures, with funding often being lobbied

for, and frequently sourced from well-connected company officials rather than being

centrally administered. Even where ‘official’ support was gained, the Government of

India generally paid only a passing interest in the results. This was especially so with

regard to the earlier trips. In the case of Charles Masson, his official duties only began

in 1834, a role he took up as a plea bargain in return for exoneration for his earlier

desertion. This picture began to change in the 1830s as the EIC looked to expand their

activities to India’s northwestern regions. Noting a shortfall in knowledge the President

of the EIC Board of Control Lord Ellenborough declared in 1829 of India’s northwest

regions ‘[w]hat we ought to have is Information. The first, the second, and the third

thing a government ought always to have is Information’.li

In the second half of the 1830s the influence of knowledgeable individuals therefore

became increasingly apparent. John Kaye records that this interest coincided with the

Governor Generalship of Lord Auckland in 1835, when the works of these adventurers

11

began ‘to be seen on the breakfast-tables of our Indian statesmen, or in their hands as

they were driven to Council.’lii

Indeed, Alexander Burnes’ published work of 1834

caught this wave of private and public interest. Burnes was courted by the London

elites, including Cabinet Ministers, the Foreign Secretary, and even enjoyed an audience

with the King and Princess Victoria.liii

Whilst attention grew, the individuals working on this area were demonstrating more

coherence as a knowledge community. In the early 1830s Elphinstone and Burnes began

a correspondence that continued up until Burnes’ death in the Kabul uprising of 1841.liv

Indeed it is clear in the works of Burnes in particular, that he owed an intellectual debt

to his mentor, and even carried with him on his first trip to Kabul and Bokhara a copy of

his Account of the Kingdom of Caubul.lv

The published works of previous travelers were

clearly consumed by their successors, with many accounts referring to the exploits of

previous explorers; the names of Elphinstone and Moorcroft often arising, as well as

other itinerant individuals. Due to the fact that his survey utilized local agents as an

information source, Elphinstone’s trail blazing also left a legacy of contacts with whom

the British were able to engage, and the exploits of feringhi (foreigners) left a lasting

impression on local actors.lvi

The most significant of those who Elphinstone met was

Shah Shuja himself whose exile from the Afghan throne during the 1820s and 1830s

was funded under a British pension.

Charles Masson and Henry Pottinger were also in regular contact throughout the 1830s

due to Pottinger’s initial financial support of Masson’s archaeological work in and

around Kabul.lvii

Pottinger’s familiarity with Masson led to his being introduced to

Claude Wade, Political Agent at Ludhiana, who negotiated his recruitment as an agent.

In 1837 Burnes finally met with Masson in Kabul at the end of Burnes’ commercial

mission up the Indus; a trip that in itself had produced a wealth of information that was

sent back to the Governor General.lviii

Between this core group of actors, we see in the archives therefore, the emergence not

just of a knowledge community, but increasingly a policy community. Whilst the

transition between these two cognitive realms is necessary to understanding how

knowledge of Afghanistan guided policy, it is to the content of this knowledge that we

turn to next.

12

Knowing Afghanistan as a political community

Whilst it is important not to create an undue sense of commonality between the works

of the early European explorers of Afghanistan there are nonetheless some similarities.

The works of Elphinstone, Masson, and Burnes, were located in the stylistic modalities

of colonial knowledge.lix

Accordingly, the material recorded in these accounts was

voluminous. This style was not simply an extreme form of empiricism, but was rooted

in a wider intellectual milieu of colonial knowledge practice. Inspired by the works of

Adam Smith surveyors sought to identify the interrelationships between all aspects of

man’s life within society: economic, political, cultural and social. The object was to

ascertain at what stage of development the society could be said to exist at (hunting;

pastoral; agricultural; or commercial), and necessitated a wealth of collected data,

including historical trajectories.lx

These methods created a generic set of interpretations

by which it was proposed ‘Asiatics’ could be judged.lxi

This led to the related

intellectual tradition of ‘philosophical’ or ‘conjectural’ history which suggested that

through the accumulation of a wealth of data, it was possible to reconstruct the

development of societies, and what is more, identify affinities with societies

elsewhere.lxii

As such, the imagined polity and society of Afghanistan was in part a

reflection of intellectual fashions, and in part the importation of familiar models of

South Asian (and Persian) societal development. Moreover, this had the effect of

infusing colonial histories of foreign territories with imported notions of legitimacy.

Through conjectural history the development of a political community was often viewed

through these accounts as the history of the rise of one elite group over another. When it

came to interacting with these communities the British were therefore crafting their own

structures of significance, interpreted through their own culturally located intellectual

understandings.

Despite these similarities important differences between these works should be

acknowledged. Personalities mattered in this sense. Whilst the rather formal

encyclopedic style of Elphinstone betrayed his more academic approach, the

travelogues of Burnes and Masson reflected their more pioneering spirit. These later

adventurers were also located within a changing cultural context that included the

13

expansion of printing culture and public knowledge in Britain itself. Burnes in particular

typified a romantic era of European exploration that included the voyages of Charles

Darwin and, later on, David Livingstone in Africa. This public thirst for adventure gave

prominence to the heroic deeds of those who had travelled to distant lands. Differences

in style also resulted from institutional contexts. Alongside their published travel

accounts Burnes and Masson’s official works betrayed the bureaucratizing and

professionalizing trends that had shaped the East India Company throughout the first

half of the nineteenth century.

Elphinstone’s Account was by far the most comprehensive and complete of these works

and his status as the foremost colonial expert on Afghan matters had a strong influence

on subsequent works. For Hopkins, this influence amounted to what he terms the

‘Elphinstonian episteme’, one that ‘definitively delineated the universe of the knowable

regarding Afghanistan. All subsequent information, recorded and archived for the

colonial state, would have to be fit into this episteme in order to be transformed into

knowledge.’lxiii

This knowledge order was both enabling and constraining; whilst it

allowed the British a structure of legibility it also limited the flexibility of their

observations.

I. Bounding Afghanistan

If as Benedict Anderson has argued, the nation is an ‘imagined community’, then we

should not be surprised at multiple, competing definitions of Afghan nationhood.lxiv

Nigel Allan identifies three. Firstly, the genealogical definition referring to the five

major Pakhtun/Pashtun tribal groups living in and around the Peshawar vale; secondly,

a wider genealogical definition referring to all Pakhtun/Pashtun groups; finally, the

territorial definition which takes ‘Afghan’ as meaning anyone who lives within the

territorial limits of the modern Afghan state.lxv

Elphinstone, facing this ambiguity, noted

that the diversity in Afghan society, government, and bodily practice made it difficult

‘to select those great features, which all possess in common, and which give a marked

national character to the whole of the Afghauns [sic]’.lxvi

Despite this, it was the first of

these definitions that he preferred, an observation described by Hopkins as the

‘Pashtunization’ of Afghanistan.lxvii

Elphinstone’s observation on the origins of the

14

Afghan nation was based in part on scriptural accounts documenting the history of the

Afghan people and written in Persian,lxviii

but it was equally clear that he saw the

Afghan polity, in a territorial sense, as a multinational entity. The Afghans co-existed

alongside Persians (including Tajiks), Balochi, ‘Tatars’, ‘Indians’, and other

‘miscellaneous tribes’.lxix

A related challenge facing explorers concerned where the borders of the Afghan polity

lay. In part this was the projection of culturally contingent understandings of national

space.lxx

In addition, cartographic practice was now a mainstay of Imperial

administrative practice - an expression of a supposedly scientific approach to

understanding the territories the British found on their frontiers, but also a familiar

language that allowed the British to converse over territories in which spatial patterns of

authority were unfamiliar to them.lxxi

This certainly applied in Afghanistan, an area

whose social formations ‘thrived on territorial flexibility, defining space relationally’,

and where ‘[d]istance, especially in regards to authority, was judged not in terms of

farsangs,lxxii

but in a genealogical idiom.’lxxiii

The cartography was further complicated

by the fact that Shah Shuja’s authority was shrinking at the time. ‘The Kingdom of

Kabul’ was in a process of flux as competing groups sought political power.

Elphinstone’s approach to overcoming this ambiguity was to adopt ‘the test made use of

by the Asiatics themselves’ defining ‘the King’s sovereignty as extending over all the

countries in which the Khootbalxxiv

is read and the money coined in his name.’lxxv

This

definition, which ignored alternative foundations of authority, created an artificially

expanded cartography of Afghanistan’s territory shown in Figure 1.

15

Figure 1: Map Taken from Elphinstone's Account Labelled 'Caubul On a Reduced Scale, Shewing [sic] its

Relative Situation to the Neighbouring Countries'. As a rough guide, the modern territorial state of

Afghanistan is bounded on its north, south, east, and west, by Bulkh (Balkh), Caubul (Kabul), Candahar

(Kandahar), and Farrah (Farah), respectively. Additional important population centres include Heraut

(Herat) in the northwest, and Peshawer (Peshawar) to the east of Kabul, on the present day border

between Pakistan and Afghanistan.

Such cartographic representations of the region appear to demonstrate a failure on the

part of the British to comprehend the reality of facts on the ground, but in truth the

British were well aware of the ambiguity of territory in the region. The frontier of the

British Empire did not end with a line on the map, but rather faded out as influence

became more tenuous, and knowledge more sparse. Indeed British diplomatic and

commercial engagement in the region, including overseeing treaty negotiations, and

mediating between warring parties can be seen as a partly a process of formalizing and

apportioning territorial limits and possessions. A process that was also underway in

India, which itself remained territorially and politically ill-defined. In the areas

inhabited by ‘Afghans’ however, British involvement was far more limited, and thus

contributed to a sense of unease with what ‘Afghanistan’ meant in territorial terms. In

16

effect, this territorial ambiguity was the manifestation of competing notions of

sovereignty between colonial and Afghan understandings, an ambiguity that disturbs the

more rigid understandings of sovereignty familiar to IR. This ambiguity was clear in the

forms of political authority that Elphinstone identified.

II. Authority, state, and tribe

For Elphinstone, royal authority in Afghanistan was the hereditary preserve of the

Saddozai sub-tribe of the Durrani tribal federation. He identified the principal function

of the government apparatus to be that of deriving revenue, the main source of which

was from land tenure, with the main expenses of the court being the payment of the

army, the household, the court establishment, and the clergy.lxxvi

In order to reduce the

tax burden on affiliated tribes, and therefore bolster the legitimacy of the King, revenue

was drawn largely from non-Afghan constituencies such as Hindus and the territories of

Peshawar and Kashmir – the loss of these two territories under Shah Shuja and Dost

Muhammad Khan respectively, would put pressure on both rulers. The army was drawn

mainly from Durrani rulers who were obliged to provide troops in return for their land

grants (Tiyuls). In this sense, as Elphinstone summarized, ‘[t]he King’s object with the

Afghaun [sic] tribes is, to get men from the western, and money from the eastern.’lxxvii

But in assessing the authority of the King over the country as a whole, Elphinstone

witnessed a tension between Royal authority, and the rule of the ‘tribes’, an observation

that highlights the sense of sovereign ambiguity that prevails in his account.

Elphinstone’s Account is perhaps most influential in the delineations he drew between

these independent tribal ‘republics’. At the centre of this was the authority ascribed to

what Elphinstone perceived as the royal clan of the Saddozai, a sub-tribe of the Popalzai

tribe which was itself part of the wider Durrani federation. For Elphinstone, the

Durranis were the Afghan political class, ‘the greatest, bravest, and most civilized in the

nation’.lxxviii

Durrani clan leaders formed the court nobility deriving ‘command and

influence from the King’s authority’,lxxix

and the patronage he disbursed.

But this regal elite coexisted within the Afghan territory with a panoply of semi-

autonomous tribal units. Whereas the colonial construct elsewhere in India had led the

17

British to view ‘tribals’ as outsiders somehow cut off from Hindu and Muslim

society,lxxx

Afghanistan presented a community based almost entirely on this concept to

which the British had attached a value judgment. Rather than seeing this pluralism as a

source of weakness however, Elphinstone praised the ‘high spirited republicanism’ of

the Afghan political community – a healthy tension in which military support from the

periphery was exchanged in return for non-interference by the ruling elite. This, he

argued, ensured defence against ‘tyrants’ and paradoxically guarded against a collective

descent into ruin across the entire country. The ‘tribes’ thus presented an alternative

conceptual lens through which to view the Afghan polity.

By the time of Masson’s arrival in Afghanistan in 1827, this ‘high spirited

republicanism’ had given way to internal disorder. The twin effects of a civil war

between the Saddozai, and the Mohammadzai clan of the Barakzai tribe, and the

expansionist moves of Ranjit Singh’s Sikh Kingdom in the Punjab, had left the rule of

Dost Muhammad Khan constrained to the immediate areas surrounding Kabul, and had

cut off important sources of external revenue. In the eyes of the British, this disorder

had opened up the state for pretenders to the throne. Masson, operating in the

ethnographic mode elaborated by Elphinstone, wrote ‘if united under a capable chief’,

the Ghilzai tribe, (who inhabited the area between Kandahar and Ghazni) ‘might … in

the present state of the country, become the most powerful’.lxxxi

Burnes’ account concurred, claiming that the ‘royal house of Cabool’ no longer existed

‘as an ostensible part of the government’ with the different chiefs ruling independently

of one another.lxxxii

Power, he claimed, was now fragmented between the population

centres of Kabul, Kandahar, Peshawar, and Herat, with Peshawar effectively a

dependency of the Sikhs, and Herat the last remaining outpost of the fallen Saddozai

dynasty.

Faced with multiple challengers for authority, including from within his own family,

Dost Muhammad Khan had opted to consolidate his rule in the territories surrounding

Kabul through a campaign of pacification and revenue collection – a process

documented in Masson’s narrative. He had also shifted the foundations of his authority

by issuing public expressions of his adherence to the Islamic scripture of the Sunna. The

purpose was twofold: firstly, this call to faith enabled his constituency base to grow,

18

taking advantage of the military role of the Ghazi (religious fighters) drawn from the

hills overlooking the Indus. Secondly, it allowed a move away from the genealogical

foundations of the previous Saddozai claim to royal privilege. Indeed Dost Muhammad

Khan was able to bridge the Sunni/Shi’a divide that Masson had noted as a growing

schism within the Afghan political community through his maternal Shi’a descent. This

growing role of religion in political affairs was noted by the British who were apt to

view unfolding events as examples of ‘fanaticism’.

In contrast to Elphinstone’s survey-based, top-down, elite-driven, concept of Afghan

political authority, based on a unitary conception of co-dependent tribal entities, Masson

presents a more fluid impression of rule. The outcome is the representation of Dost

Muhammad Khan as primus inter pares, negotiating a balancing act between the co-

opting of potentially useful community leaders, whilst keeping them at a sufficient level

of political power that they wouldn’t pose a threat to his rule. This was the

demonstration of a negotiated process of rule that Elphinstone had hinted at in his

Account, but he was perhaps less exposed to it in practice through his engagement with

mainly courtly elites and their cohorts – a group he referred to as ‘the great’. Burnes and

Masson on the other hand were both courted by competing elites, giving them a more

fragile impression of Afghan rulership.lxxxiii

The cumulative knowledge base upon which British conceptions of Afghanistan were

based was therefore conceptually coherent but factually fractured. Observers such as

Masson and Burnes found their observations frequently evaded a simple description. On

the one hand this state of affairs reflected the ongoing ructions within the Afghan

political classes at the time of their visits, but in a wider sense it also represented the

uncertainties of what was a process of transculturation. The formation of the

Afghanistan knowledge community was not simply a process of information gathering,

but was also a process in which political conceptual criteria were transferred into an

Afghan political universe, including notions of territory and sovereign authority. This

was bound to generate uncertainties as colonial and Afghan understandings entered a

process of dialogue. Increasingly however, policy imperatives drove a process of

closure over the ambiguity that inhered within this construct.

19

Knowing Afghanistan at the contact zones: a policy science for the powerful

As previously discussed, the 1830s saw a growing policy interest in affairs concerning

the northwest of India. Not only did this prompt a greater demand for knowledge but it

also increasingly attached such knowledge to policy prescriptions. The institutional

structure within which this process took place can be understood by reference to what

Ulrike Hillemann terms ‘contact zones’: sites for the transference of information,

knowledge, concepts and ideas between the periphery and the metropole, and back

again.lxxxiv

Three sites were central to the picture that the British were beginning to build

up. Firstly, London, the Cabinet, and the Court of Directors of the East India Company

(EIC); secondly Calcutta, and the office of the Governor-General; and thirdly the

Punjab, and the political officers. Important to this understanding is that in London the

EIC was reaching the twilight years of its role in determining political concerns in India.

Despite the appearance of metropolitan centralism (The Prime Minister still retained the

power to declare war, for example), British India was beginning to craft a distinct

foreign policy for itself, one that would be more cognizant of regional realities, rather

than as a tool of balancing what were essentially European geopolitical concerns.lxxxv

This in turn meant that the post of the Governor-General was beginning to enjoy a more

exalted position in the British colonial policy-making hierarchy; in addition the office of

the Governor General was increasingly overseeing the collection of political, military,

and diplomatic intelligence.lxxxvi

Whilst this had the effect of decentralizing the

bureaucratic process away from institutions of the EIC, it also had the effect of

recentralizing knowledge procurement within the circle of the Governor-General’s

advisors. The growing activities of the political agents in the northwest were in part a

reflection of this and their engagement with figures such as Charles Masson had the

effect of officialising what had previously been a more entrepreneurial and open-ended

knowledge community.

Driving this at the metropole was an intellectual and policy climate that since the late

1820s had exhibited signs of growing ‘Russophobia’. Two works by Colonel De Lacy

Evans – On the Designs of Russia (1828), and Practicability of an Invasion of British

India (1829) - had captured the imagination of a public wary of Russian moves against

20

the Ottoman Empire, and copies were duly sent to the British representatives in Persia

and Bombay.lxxxvii

In Calcutta however, the assessment of the nature of this threat

differed: the impact of ‘rumour’ amongst the local population was considered more

dangerous. Believing that their position in India owed much to the perception of their

superior military power the British worried constantly about any threat to this fragile

edifice, particularly at the frontiers of their territory. As Lord Ellenborough, one of the

more paranoid policy makers noted in 1830: ‘We dread … not so much actual invasion

by Russia, as the moral effect which would be produced amongst our own subjects and

among the Princes with whom we are allied, by the continued apprehension of that

event.’lxxxviii

The principal method of counteracting Russian influence had been the steady expansion

of trade and these efforts centred on the Indus River. It was envisaged, that by

establishing an ‘entrepôt’ of trade on the river, the benefits would emanate along

commercial networks that worked their way westward through the Gomal Pass towards

Herat, and northwards through the Bolan and Khyber passes, to Kabul and beyond. It

was also hoped that this trading hub would present an alternative to the Russian-

controlled trade fair of Niznii Novgorod the north, attracting merchants from Central

Asia and thereby reducing the commercial influence of Russia. This was more than

simply trade ‘following the flag’. Commercial enterprise was very much the ‘idiom of

British governmentality’ at this time.lxxxix

The exploits of Colonel Chesney in

Mesopotamia provided a model for the project and the unsuitability of the Euphrates

and Tigris meant the possibility of steamboats being transferred southwards; a new

technology that it was hoped would benefit communications, trade, and security across

the British territories.xc

To this end, in 1837 Alexander Burnes, along with three EIC

officials, was sent on a second surveying mission to assess the commercial viability of

the Indus River and the prospects of trade with Kabul. Burnes’ trip therefore satisfied

the dual purposes of a commercial urge, and a desire for accurate political intelligence.

The result was a far more policy-driven development of existing knowledge, the output

of which went directly to the Governor-General, who as late as 1837 was lamenting the

‘haze of confusion’ that existed beyond Lahore.xci

The perception of a fractured Barakzai polity in the northwest had encouraged a spirit of

cautious engagement with Afghan affairs through much of the 1830s and the wisdom of

21

commercial enterprise was not without controversy.xcii

Dost Muhammad Khan faced not

only external threats from the Sikhs, but internal threats too from his Barakzai brothers

at Kandahar and Peshawar. Aside from a brief alliance with Kabul during Shah Shuja’s

failed attempt to reclaim his throne in 1833-4, Kandahar was autonomous, whilst

Peshawar – although ruled by Dost Muhammad Khan’s brother - was perceived to be

firmly under the yoke of Sikh power. Moreover, Kandahar’s geographic position left it

vulnerable in British eyes to Persia, which was at the time laying siege to Herat. In a

variant on ‘domino theory’, it was feared that should Herat fall, Kandahar would be

forced to align with the Persians; and that in this scenario, the Persian community,xciii

including Persian elements of the military – the Qizilbash - in Kabul would be

encouraged into insurrection against Dost Muhammad Khan. Behind Persia, it was

believed, lay Russian support.

This regional threat complex is significant. On the one hand it shows the manner in

which quasi-causal stories were beginning to emerge, but more significantly for our

purposes here, it demonstrates the ways in which the British were socializing their own

geopolitical fears into their knowledge-based lexicon of regional politics. The

perception of threat heightened certain features of the Afghan polity, as the British

imagined it, including inter-Barakzai feuds. Dost Muhammad Khan’s maternal Shi’a

descent thereby became a threat, as Burnes argued, because it potentially aligned him

with the Shi’a state of Persia.xciv

The significance of this was not lost on Governor

General Auckland who in June 1838 described Dost Muhammad Khan as ‘a

kuzilbashxcv

[sic] to the Westward’.xcvi

It is notable that those aspects of Elphinstone’s

argument that drop out include the positive tension between centre and periphery that

had previously held together the rule of Shah Shuja. British preconceptions also entailed

a revised view of the political geography. Herat became viewed as the bastion of

defence against Persian expansionism, despite the schism that existed by virtue of the

ruler of Herat being a Saddozai and therefore outside of the ruling family of Dost

Muhammad Khan.

Sovereign competence: Dost Muhammad Khan and Shah Shuja

22

Having shifted his foundations for sovereign legitimacy from lineage to Islam, Dost

Muhammad Khan came under particular scrutiny in his capacity to govern. Burnes

noted that shrinking land revenue had been balanced by closer sovereign oversight of

the trading customs house and an increased, but not overbearing commercial tax. Within

Kabul, his esteem was judged as high. Despite this however, Burnes lamented the

continual frittering away of this revenue on costly military campaigns. The Afghan-Sikh

battle of Jamrud in early 1837 had underlined this point, and as Burnes argued ‘a

diminution of [Dost Mohammed Khan’s] enemies will have the same effect as an actual

increase to his resources.’xcvii

The outcome of this was uncertain however. Burnes noted

that even if Afghan-Sikh rivalry could be negotiated through an agreement over

Peshawar (and this was considered unlikely), the ‘Mahomedan tribes inhabiting the

mountains of Eastern Afghanistan … who now regard the Ruler of Cabool as the

Champion of Islam, might then view him simply as an ambitious ruler seeking for

personal aggrandizement’.xcviii

In addition, the call to faith had alienated the Shi’a

community of the Persian Qizilbash who bolstered his authority. Dost Muhammad

Khan’s shift in the foundations of his legitimacy was therefore not only unfamiliar to

the British, but - based on their increasingly nucleated view of the Afghan political

community - ultimately unsustainable.

The British retained an ideal-type image of the Afghan polity under the former Saddozai

monarchy – and this was sustained by the intellectual heritage of Elphinstone.

Moreover, Shah Shuja’s exile in the Punjab had allowed a regular channel of

communication between himself and the Governor General, which he used to encourage

the British of his continued popularity with vast swathes of the Afghan political

community, as well as to encourage a perception of the illegitimacy of the Barakzai.xcix

This impression was particularly strong at Calcutta. The appeal, to the British, of a

return to Saddozai rule was not just a result of the intellectual legacy of Elphinstone’s

concept of traditional authority, or in Shuja’s continued assertions of his legitimacy, but

also in a sense of security that the British derived from the concept of a unitary political

formation. By September 1837, the Governor General Lord Auckland was referring to

the ‘triple power’ of Afghanistan.c For the British, Afghanistan increasingly resembled

a fragmented polity amidst a sea of predatory tribal chieftaincies, monarchies, and

states. Shah Shuja offered a sovereign order that cohered with the outdated official

vision of a stable Afghan polity.

23

In March of 1838, the Secretary to the Governor General, William Hay Macnaghten,

wrote to Burnes and Masson asking for proposed measures to ‘counteract’ the policy of

Dost Muhammad Khan. Both replied with the suggestion that a British backed venture

led by Shah Shuja with limited British financial and military support would achieve

British ends. Burnes opined that British policy should be to make Kabul ‘as strong as

we can make it and not weaken it by divided power … we should consolidate the

Afghan power west of the Indus and have a King and not a collection of Chiefs.’ He

added: ‘“Divide et impera” is a temporising creed at any time and if the Afghans are

united we … bid defiance to Persia and instead of distant relations we have every thing

under our own eye and a steadily progressing influence all along the Indus.’ci Masson

concurred: ‘In aiding the restoration of Shah Sujah’, he proposed, ‘the British

Government would consult the feelings of the Afghan nation, among which his

popularity is great, and who even wonder that the Government has not before done it. If

he avowedly advanced under British auspices, his success would be prompt and certain,

little or no blood would be shed – he would be joined by all who are discontented with

the Barrakzai rule … Even the Powerful Kazzalbash [sic] faction at Kabul would

acknowledge the Shah, for they have no other object in promoting and abetting the

designs of Persia, than to rid themselves of Dost Mahomed Khan.’ Masson further

argued that the strength of this proposal lay in the potential it had to unite Kabul with

Herat, which remained the last outpost of Saddozai rule under Shah Kamran.cii

At the time Macnaghten made his request of Masson and Burnes he was on his way to

meet with Ranjit Singh to discuss the possible options with regards to dealing with Dost

Muhammad Khan. The Anglo-Sikh alliance was judged more important than placating

the needs of the Barakzai ruler of Afghanistan, and during the summer months the

Governor General and his cohorts developed the plans for a military invasion of

Afghanistan that would replace Dost Muhammad Khan with Shah Shuja. At this point,

the advice of the knowledge community became subsumed by the policy, but the

framing they provided had nonetheless, guided that policy. At this moment, the

knowledge formed by the Afghanistan knowledge community was marking a final shift

from the ‘embodied knowledge’ that had been the mainstay of the early Company state,

to the more official, or ‘institutional knowledge’ of the colonial state.ciii

This entailed a

24

more fixed conceptual order, as the ambiguities of the works of the early European

explorers were ironed out for reasons of policy expediency.

It remains to be said that of Elphinstone, Masson, and Burnes, none agreed with the

eventual policy path chosen. Masson and Burnes both claimed that their advice had

been misinterpreted, and that they had only proposed a light financial and military

backing to Shah Shuja, in order to tip the balance of threat in his favour, and win over

prevaricating groups within Afghanistan’s political community.civ

Masson resigned

from Government service in 1840 after being wrongfully detained as a spy for the

Afghans in Quetta. In 1839 Elphinstone wrote to Burnes, who was by that time envoy at

Kabul, to offer his opinions on the policy. ‘I have no doubt you will take Candahar and

Cabul and set up Shuja but for maintaining him, in a poor, cold, strong and remote

country, among a turbulent people like the Afghans, I own it seems to me to be

hopeless.’cv

Masson was even more outspoken. ‘It is to be hoped’, he wrote, ‘that the

good sense of the British nation will never again permit such expeditions as the one

beyond the Indus, to be concerted with levity, and to be conducted with recklessness’.cvi

Such comments are priceless for those who seek to draw a line between the

contemporary era and the past. But as Priya Satia has urged, Marx’s famous dictum that

‘history repeats itself, first as tragedy, then as farce’, was not meant to be taken literally.

Rather, history develops dialectically, and the conditions of possibility that allow this to

happen are as much epistemological as they are material.cvii

Conclusion

In the lead up to the First Anglo-Afghan War, the British may have seen a nation in

Afghanistan, but it was a faint construct, contested both locally and regionally, and

deriving more from a need to categorize a heterogeneous community than from any

sense of the term ‘nation’ in its contemporary meaning. Moreover, the British did not

yet see a state, and where they did, it was only as the crumbling ruins of the former

Saddozai regime – a direct consequence of the impact of Elphinstone’s work.

Knowledge of Afghanistan did not cause the Anglo-Afghan War of 1838-42, but

insufficient attention to the construction of the idea of the Afghan polity in the minds of

the British has overlooked this important conditioning factor in the decision to go to

25

war. As Christine Noelle puts it, ‘[w]hile greater themes like ‘imperialism’ … point to

the origins of British action, they fail to account for the manner in which the British

attempted to extend their influence in Afghanistan’.cviii

In short, knowledge of

Afghanistan made the war imaginable in the first place.

This article has sought to demonstrate the value of attention to colonial knowledge,

which hitherto has received little attention within the IR discipline, despite the ‘imperial

turn’ of the past decade. Colonial knowledge can be viewed as a ‘register’ of imperial

international thought,cix

highlighting how core categories familiar to IR, including the

state, sovereignty, and territoriality, can be viewed as emergent phenomena, rather than

as pre-social, essentialized categories. From an IR perspective, colonial knowledge

provides not just a site for the construction of state identity by an external actor, but a

site for dialogue between competing idioms of political order – in this case a British

conception and an Afghan conception – albeit a dialogue dominated by the categories of

the former over the latter.

The development of what has been termed a ‘knowledge community’ on Afghanistan

was a manifestation of an evolving approach to the collection and use of colonial

knowledge more widely by the colonial state. In the case of Afghanistan, the use of this

knowledge in framing the policies that led to the First Anglo-Afghan War, shut down

the latitude and fluidity that was previously apparent in existing understandings of the

Afghan polity, including competing sovereign orders, and the vaguaries of

Afghanistan’s territorial reach. The knowledge community provided two principal ways

through which the British could ‘imagine’ Afghanistan. On the one hand it provided

evidence for a fractured, acephalous polity, at war with itself, unbounded, volatile and

inadequately led by a ‘chief’, who had usurped the legitimate ruling dynasty and now

threatened neighbouring powers in order to compensate for his own internal weakened

rulership. This picture was heightened by the contrasting features of the polities – even

‘states’ – surrounding Afghanistan that were apparently more bounded, stable, unified,

monarchical authorities with whom the British could more easily conduct diplomatic

and commercial exchanges. The difficulties that the British faced in scrutinizing and

categorizing Afghanistan - ‘a country which’, as Burnes described it, ‘seemed as it were

not’ - meant that the territory became more of a locational concept, continually

contested in the imperial mind.cx

The march to war in 1838 forced a process of closure

26

over this definition. The second contribution made by the knowledge community was in

providing an alternative scenario: reinstating the deposed Saddozai authority, under

Shah Shuja, who could rectify the failings of the collapsed polity the British now

believed they were seeing beyond the Indus river.

Ultimately the knowledge community that grew up around the policy problem of British

India’s northwest provided the sources for justifying a policy of conquest, and left an

intellectual legacy that outlasted its relevance. The failure of the project to consolidate

British presence in Afghanistan led to a sustained period of exclusion as Afghanistan

settled into the colonial imagination as the epitome of a terra incognita, a true ‘outlaw

state’.cxi

The closest contact that the British had with the Afghans during this period was

through the tribes of the North West Frontier and a tenuous system of ‘native

informants’. The form of administration here mirrored the sense of exclusion that the

British felt towards the territories to the north and west: not seeking to ‘integrate,

civilize and modernize’, but rather to ‘contain, conserve and traditionalize’, drawing

‘deeply upon the well of colonial memory’, thereby preserving the works of Elphinstone

et al.cxii

Indeed Elphinstone’s work, being the most complete of its kind, had the effect

of freezing in time British conceptualizations of the Afghan territories long after its

relevance had expired. As late as 1887, over 70 years on from the first publication of

Elphinstone’s Account, Blackwood’s Edinburgh Magazine was still citing it as a key

text.cxiii

Of Masson’s works, the famous nineteenth century geographer Sir Thomas

Holdich wrote, ‘the most amazing feature of Masson’s tales of travel is that in all

essential features we knew little more about the country of the Afghans after the second

war with Afghanistan than he could have told us before the first.’cxiv

Colonial

knowledge on Afghanistan has cast a long shadow.

Rescuing Afghanistan’s imperial encounter from the oblivion of great power diplomatic

history demands closer attention to the sunken histories of this period. Imperial histories

have for some time moved on from the sterile categories of official practice and instead

embraced social, cultural and intellectual history to thicken their narratives and to give

voice to the voiceless. Whilst the ‘imperial turn’ in IR has created space for the

incorporation of such work into the discipline, there is much left to explore. In arguing

for greater attention to colonial knowledge in particular this article has sought to

conduct IR in the imperial mode, in a way that does not do damage to either the history,

27

or to the insights that IR can bring to the table. Doing so requires shedding traditional

disciplinary adherence to unchanging conceptual baggage.

28

i. Rory Stewart, The Places in Between (London: Picador, 2005). ii. Manan Ahmed, ‘Adam’s Mirror: The Frontier in the Imperial Imagination’, Economic and Political Weekly, XLVI:13 (2011); Benjamin Hopkins and Magnus Marsden, Fragments of the Afghan Frontier (London: Hurst, 2011). iii. In a documentary broadcast on BBC2 in 2012 Stewart described one of these men, Alexander Burnes, as the greatest diplomat Britain had ever seen. ‘Afghanistan: The Great Game – A Personal View by Rory Stewart’, BBC2 (Broadcast on 28 May 2012). iv. Stewart was also courted by Richard Holbrooke following the latter’s arrival in the post of ‘AfPak’ special envoy. Rory Stewart, ‘Midnight Moments: Exposing the truth and taking full responsibility for Afghanistan’, Huffington Post (20 December 2010), available at: {http://www.huffingtonpost.com/rory-stewart/post_1457_b_799063.html} accessed 19 October 2012}. For a critique of Foreign Policy ‘expertise’ see: Manan Ahmed, ‘Flying Blind: US foreign policy’s lack of expertise’, The National (4 March 2011), available at: {http://www.thenational.ae/arts-culture/books/flying-blind-us-foreign-policys-lack-of-expertise} accessed 19 October 2012}. v. Peter Hopkirk, The Great Game (London: John Murray, 2006); John Ure, Shooting Leave (London: Constable, 2009). vi. Benjamin Hopkins, The Making of Modern Afghanistan (Cambridge: CUP, 2006), p. 34. See also, Benjamin Hopkins, ‘The Myth of the ‘Great Game’: The Anglo-Sikh Alliance and Rivalry’, University of Cambridge Centre of South Asian Studies Occasional Paper, No.5 (2004). vii. In this sense the article builds on Hopkins’ important contribution to the modern literature The Making of Modern Afghanistan, in which he refers to the Afghan ‘proto-state’. This terminology has implications for approaches pertaining to the International Relations discipline and is worthy of further interrogation. viii. John M. Hobson, ‘What’s at stake in ‘bringing historical sociology back into international relations’? Transcending ‘chronofetishism’ and ‘tempocentrism’ in international relations’, in Stephen Hobden and John M. Hobson (eds), Historical Sociology of International Relations (Cambridge: CUP, 2002), pp. 3-41. ix. Hopkins, The Making of Modern Afghanistan; Benjamin Hopkins and Magnus Marsden, Fragments of the Afghan Frontier (London: Hurst, 2011); Christine Noelle, State and Tribe in Nineteenth-Century Afghanistan: The Reign of Amir Dost Muhammad Khan (1826-1863) (Richmond, Surrey: Curzon Press, 1997); Robert Johnson, The Afghan Way of War (London: Hurst, 2010); Shah Mahmud Hanifi, Connecting Histories in Afghanistan: Market Relations and State Formation on a Colonial Frontier (Stanford: Stanford University Press, 2011).

29

x. Stephen H. Haber, David M. Kennedy, and Stephen D. Krasner, ‘Brothers under the Skin: Diplomatic History and International Relations, International Security, 22:1 (1997), pp. 34-43. xi. Nico Stehr and Volker Meja (eds), Society and Knowledge, 2nd Edition (New Brunswick, NJ: Transaction Publishers, 2005), p. 13. xii. Stehr and Meja (eds), Society and Knowledge, 2nd Edition, p. 13. See also Peter L. Berger and Thomas Luckmann, The Social Construction of Reality (London: The Penguin Press, 1967); Peter Burke, A Social History of Knowledge (Cambridge: Polity Press, 2002). xiii. Colin Elman and Miriam Fendius Elman (eds), Bridges and Boundaries: Historians, Political Scientists, and the Study of International Relations (London: The MIT Press, 2001); George Lawson, ‘The Eternal Divide? History and International Relations’, The European Journal of International Relations, 18:3 (2012), pp. 203-26; David Armitage, ‘The Fifty Years’ Rift: Intellectual History and International Relations’, Modern Intellectual History, 1:1 (2004), pp. 97-109. xiv. See for example, Michael Cox, Tim Booth, and Ken Dunne, ‘Empires, Systems, and States: Great transformations in international politics’, Review of International Studies, 27:1 (2001), pp. 1-15. xv. Christian Reus-Smit, ‘Reading History through Constructivist Eyes’, Millennium 37:2 (2008), p. 395. xvi. George Lawson, ‘The Promise of Historical Sociology in International Relations’, International Studies Review, 8:3 (2006), 404. See also, Lawson, ‘The Eternal Divide?’. xvii. John Hobson, George Lawson, and Justin Rosenberg, ‘Historical Sociology’, LSE Research Online (June 2010), available at: {http://eprints.lse.ac.uk/28016} accessed 18 July 2012}. xviii. Reus-Smit, ‘Reading History through Constructivist Eyes’. xix. Quoted in Daniel Jacobi, ‘On the ‘Construction’ of Knowledge and the Knowledge of ‘Construction’’, International Political Sociology, 5:1 (2011), pp. 94-7. xx. Lawson, ‘The Eternal Divide?’, pp. 209-10. xxi. Lawson, ‘The Eternal Divide’, p. 205. xxii. Tarak Barkawi and Mark Laffey, ‘Retrieving the Imperial: Empire and International Relations’, Millennium, 31:1 (2002), pp. 109-27; Robert Vitalis, ‘Birth of a discipline’, in David Long and Brian Schmidt (eds), Imperialism and Internationalism in the Discipline of International Relations (Albany, NY: State University of New York Press, 2005), pp. 159-82.

30

xxiii. Tarak Barkawi, ‘Empire and Order in International Relations and Security Studies’ in Robert A. Denemark (ed), The International Studies Encyclopedia, Vol. III (Chichester: Wiley- Blackwell, 2010), pp. 1360-1379. See also David Long and Brian Schmidt (eds), Imperialism and Internationalism in the Discipline of International Relations (Albany, NY: State University of New York Press, 2005). xxiv. John Agnew, ‘The Territorial Trap: The Geographical Assumptions of International Relations Theory’, Review of International Political Economy, 1:1 (1994), pp. 53-80. xxv. Tarak Barkawi and Mark Laffey, ‘Retrieving the Imperial: Empire and International Relations’, Millennium, 31:1 (2002), pp. 109-27; Martin Shaw, ‘Post-Imperial and Quasi-Imperial: State and Empire in the Global Era, Millennium, 31:2 (2002), pp. 327-336. xxvi. Jordan Branch, ‘’Colonial reflection’ and territoriality: The peripheral origins of sovereign statehood’, European Journal of International Relations, 18:2 (2012), pp. 277-97. xxvii. Edward Keene, Beyond the Anarchical Society: Grotius, Colonialism and Order in World Politics (Cambridge: Cambridge University Press, 2002). xxviii. Gerry Simpson, Great Powers and Outlaw States (Cambridge: CUP, 2004). xxix. Johnson, The Afghan Way of War, 2. xxx. Thomas Barfield, Afghanistan: A Cultural and Political History (Oxford: Princeton University Press, 2010). xxxi. Noelle, State and Tribe in Nineteenth-Century Afghanistan, p. 38. xxxii. On the relevance of contextualism to IR see Duncan Bell, ‘The Cambridge School and World Politics’, The Global Site (2001), available at: {http://www.theglobalsite.ac.uk/press/103bell.htm} accessed 19 October 2012}. xxxiii. On the concept of ‘closure’ and the sociology of knowledge see Jan Golinski, Making Natural Knowledge: Constructivism and the History of Science (Cambridge: CUP, 1998). xxxiv. Where ‘information’ refers to that which ‘is relatively ‘raw’, specific and practical’. Burke, A Social History of Knowledge, pp. 11-12. xxxv. Berger and Luckmann, The Social Construction of Reality, p. 27, p. 30. xxxvi. Nico Stehr and Volker Meja (eds.), Society and Knowledge, 2nd Edition (New Brunswick, NJ: Transaction Publishers, 2005), p. 13.

31

xxxvii. Stehr and Meja (eds.), Society and Knowledge. xxxviii. Bernard S. Cohn, Colonialism and Its Forms of Knowledge (Princeton: Princeton University Press, 1996), p. xi. (Quoting the foreword by Nicholas Dirks). xxxix. Cohn, Colonialism and Its Forms of Knowledge. xl. Robert A. Stafford, ‘Scientific Exploration and Empire’ in Andrew Porter and Alaine Low (eds), The Oxford History of the British Empire: Volume III, The Nineteenth Century (Oxford: OUP, 1999), p. 311; Edward Said, Orientalism, (London: Penguin Books, 2003). xli. James C. Scott, Seeing Like a State, (Yale: Yale University Press, 1999), p. 11. xlii. Christopher A. Bayly, Empire and Information (Cambridge: CUP, 1996); Eugene Irschick, Dialogue and History: Constructing South India , 1795-1895 (Berkley, LA: University of California Press); Mary Louise Pratt, Imperial Eyes: Travel Writing and Transculturation (Abingdon: Routledge, 2008); Peter Burroughs, ‘Imperial Institutions and the Government of Empire’, in Andre Porter and Alaine Low (eds), The Oxford History of the British Empire: Volume III, The Nineteenth Century, pp. 170-97; Ulrike Hillemann, Asian Empire and British Knowledge (Cambridge: CUP, 2009). xliii. Hopkins, The Making of Modern Afghanistan, p. 8. xliv. This process of centre-periphery exchange was visible in the construction of Afghan national identity prior to the arrival of the British. Nile Green has recently shown how Mughal-era diasporic networks had already fostered a sense of ‘Afghan’ identity amongst Pashtun elites in advance of British influence in the region. Nile Green, ‘Tribe, Diaspora, and Statehood in Afghan History’, The Journal of Asian Studies, 67:1 (2008), pp. 171-211. xlv. Hopkins, ‘The Myth of the ‘Great Game’. xlvi. Mountstuart Elphinstone should not be confused with his relative General William Elphinstone who led the disastrous retreat from Kabul during the First Anglo Afghan War. xlvii. Mountstuart Elphinstone, An Account of the Kingdom of Caubul, and its Dependencies in Persia, Tartary and India, Second Edition [in two volumes] (London: Richard Bentley, 1819). xlviii. Masson’s real name was in fact James Lewis. ‘Charles Masson’ was a pseudonym invented in order to cover for his desertion. Gordon Whitteridge, Charles Masson of Afghanistan (Warminster: Aris and Phillips Ltd, 1986); Bijan Omrani, ‘Charles Masson of Afghanistan: Deserter, Scholar, Spy’, Asian Affairs, 39:2 (2008).

32

xlix. Henry Pottinger’s nephew, Lieutenant Eldred Pottinger, also played a prominent role later on in the defence of Herat during the Persian siege of 1837. l. Claude Wade also used his position as a conduit for intelligence reports from Kabul to the Governor General to warp the information contained within them in a manner that was more favourable to his preferred candidate for the throne, Shah Shuja. J.A. Norris, The First Afghan War 1838-1842 (Cambridge: University Press, 1967), p. 139; Dalrymple, Return of a King, p. 65. li. Whitteridge, Charles Masson of Afghanistan, p. 18. lii. John William Kaye, History of the War in Afghanistan, Volume I (London: Richard Bentley, 1857). liii. Kaye, History of the War in Afghanistan, Volume I, p. 179. See also, India Office Records (hereafter IOR), Masson Papers, Mss Eur E.161/633, Correspondence III, 3, enclosure 2; Mss Eur E.161/6-7 (microfilm), 6a, Pottinger to Masson, July 27, 1834. Burnes’ movements during his second commercial mission to Kabul in 1837 were closely followed in the Delhi Gazette for the consumption of the British settlers in Delhi. IOR (microfilm), The Delhi Gazette, 1837-8, SM 52. liv. See IOR, Elphinstone Papers, Mss Eur F88/81, 85, 91, 105, 111. lv. Alexander Burnes, Travels into Bokhara, Volume I (London: John Murray, 1834), p. 162. Masson also requested a copy from Pottinger in 1833: IOR, Masson Papers, 20876 Mss Eur E.161/6-7/1a (microfilm). lvi. One of these agents, Mullah Najib hosted Burnes on his later trip through Peshawar. Najib was at the time on a British pension. Burnes, Travels into Bokhara, Volume I, p. 105. lvii. Omrani, ‘Charles Masson of Afghanistan’. Masson’s correspondence with Pottinger and Wade is in the Masson Papers, IOR Mss Eur.E.161/631-632. lviii. IOR/V/27/270/7, ‘Reports and Papers, Political, Geographical, and Commercial. Submitted to Government, by Sir Alexander Burnes; Lieutenant Leech; Doctor Lord; and Lieutenant Wood’. lix. Cohn, Colonialism and Its Forms of Knowledge. lx. Jane Rendell, ‘Scottish Orientalism: from Robertson to James Mill’, The Historical Journal, 25:1 (1982), pp. 43-69. lxi. Said, Orientalism. lxii. Rendell, ‘Scottish Orientalism’. lxiii. Hopkins, The Making of Modern Afghanistan, p. 17.

33

lxiv. Benedict Anderson, Imagined Communities (London: Verson, 2006). lxv. Nigel Allan, ‘Defining Place and People in Afghanistan’, Post-Soviet Geography and Economics, 42:8 (2001), p. 548. lxvi. Elphinstone, Account, Vol I, p. 238. lxvii. Hopkins, The Making of Modern Afghanistan, p. 23. lxviii. This included consultation of the genealogical histories of the founding father of the Afghan nation, Qais Abdur Rashid. Elphinstone, Account, Vol I, p. 253; p. 398. lxix. Elphinstone, Account, Vol I, p. 138. lxx. Hopkins, The Making of Modern Afghanistan. lxxi. Branch, “Colonial Reflection’ and territoriality’. lxxii. A Persian unit of measurement based on how far a man can walk in one day. lxxiii. Hopkins, The Making of Modern Afghanistan, p. 25. lxxiv. As per the original: ‘The Khootba is part of the Mahommedan service, in which the king of the country is prayed for. Inserting a prince’s name in the Khootba, and inscribing it on the current coin, are reckoned in the East the most certain acknowledgement of sovereignty.’ Elphinstone, Account, Vol I, p. 138. lxxv. Elphinstone, Account, Vol I, p. 138. lxxvi. Elphinstone Account, Vol II, p. 269. lxxvii. Elphinstone, Account, Vol II, p. 247. lxxviii. Elphinstone, Account, Vol I, p. 85. lxxix. Elphinstone, Account, Vol I, p. 86. lxxx. Christopher A. Bayly, ‘Knowing the Country’, Modern Asian Studies, 27:1 (1993), pp. 3-43. lxxxi. Charles Masson, Narrative of Various Journeys in Balochistan, Afghanistan, and the Panjab, Vol II (London: Richard Bentley, 1842), p. 204. lxxxii. Burnes, Travels, Vol. III, p. 250.

34

lxxxiii. Burnes, for example, was courted by Sirdar Sultan Muhammad Khan of Peshawar; Masson by Haji Khan, chief of Bamiyan (northwest of Kabul) and one of Dost Muhammad Khan’s principal advisers. Burnes, Travels, Vol III, p. 255; Masson, Narrative, Vol II, p. 360. lxxxiv. Hillemann, Asian Empire and British Knowledge. lxxxv. As the President of the Board of Control Lord Ellenborough proposed in late 1829 ‘the Indian Government should be authorized to act as an Asiatic power, ignoring the effect of its actions on Britain and Europe, if the Russians moved towards Kabul.’ Norris, The First Afghan War, p. 31. lxxxvi. Bayly, Empire and Information, pp. 144-5. lxxxvii. Norris, The First Afghan War, p. 30. lxxxviii. Whitteridge, Charles Masson of Afghanistan, p. 31. lxxxix. Hopkins, The Making of Modern Afghanistan, p. 38. xc. The findings of a Select Committee on Steam Communications were reported in the Delhi Gazette, IOR (Microfilm) SM52, January, 1837. See also, IOR, Broughton Papers, Add MS 36473, pp. 64, 80, 85-6, 188, xci. IOR, Broughton Papers, Add MS 36473, p. 120. For the reports see IOR/V/27/270/7, ‘Reports and Papers, Political, Geographical, and Commercial. Submitted to Government, by Sir Alexander Burnes; Lieutenant Leech; Doctor Lord; and Lieutenant Wood’. xcii. Henry St George Tucker, who was Chairman of the Company Board at the time, declined to support this second mission ‘feeling perfectly assured that it must soon degenerate into a political agency, and that we should as a necessary consequence be involved in all the entanglement of Afghan politics.’ Kaye, History of the War in Afghanistan, Volume I, p. 181. xciii. The Qizilbash are to be distinguished from the ethnic Hazara community inhabiting the central highlands of the Hindu Kush who were also of the Shi’a sect of Islam and accordingly were viewed by the British as potentially aligning with Persia in the event of a Persian invasion. xciv. IOR/V/27/270/7, p. 10. xcv. The Qizilbash were a Persian unit of the Afghan army that functioned as the personal bodyguards of the ruler of Kabul, having been originally established under the rule of Ahmad Shah (r.1747-72). xcvi. IOR, Broughton Papers, Add MS 36473, p. 262.

35

xcvii. IOR/V/27/270/7, p. 15. xcviii. IOR/V/27/270/7, p. 15. xcix. IOR/L/PS/20/MEMO1/15/3, pp. 4-5, p. 7. Shah Shuja also made three attempts to regain his throne, each of which failed. The first attempt via Kashmir in 1815, an aborted attempt in 1818, and then again 1834. The latter two had been followed by British officials and in the case of the 1834 expedition, supported by British financing. Despite this there was a reluctance to become more involved in these efforts and what was seen as an opaque political contest. William Dalrymple, Return of a King (London: Bloomsbury, 2013), pp. 36-8, 45-6, 66-73. c. IOR, Broughton Papers, Add MS 36473, p. 188. ci. IOR, Elphinstone Papers, Mss Eur F88/105, p. 46-9. cii. IOR, Broughton Papers, Add MS 36473, pp. 370-7 ciii. For this distinction see: Bayly, Empire and Information, p. 144. civ. Masson, Narrative, Vol 1, vi-viii. Burnes made a similar argument in response to a letter from Elphinstone disputing the wisdom of the invasion. In his words, ‘I never doubted we could place the Shah on the throne but that I viewed the Army as far too large – Indeed the passing of so many British troops into Afghanistan has been the prime cause of Shah Shooja’s partial unpopularity’. IOR, Elphinstone Papers, Mss Eur F88/111, p. 81. cv. IOR, Elphinstone Papers, Mss Eur F88/111, p. 28. cvi. Masson, Narrative, Vol I, p. xi. cvii. Satia, Spies in Arabia, p. 337. cviii. Noelle, State and Tribe in Nineteenth-Century Afghanistan, p. 40. cix. Duncan Bell, ‘Empire and International Relations in Victorian Political Thought’, The Historical Journal 49:1 (2006), p. 283. cx. IOR, Broughton Papers, Mss Eur F213/89, 4 July 1840, ‘Observations on the restored Government of Shah Shooja’. cxi. Simpson, Great Powers and Outlaw States. cxii. Hopkins and Marsden, Fragments of the Afghan Frontier, p. 63. cxiii. ‘In the Heart of Afghanistan’, Blackwood’s Edinburgh Magazine, January (1887), p. 81.

36

cxiv. The Second Anglo-Afghan War took place between 1878-80. Whitteridge, Charles Masson of Afghanistan, p. 41.